The relationship between lithium intake and health is a topic of minor interest, in that no-one is going to be building a rejuvenation therapy on the basis of the mechanisms by which lithium may very modestly slow aging in short-lived species. There is some evidence for greater human life expectancy to occur in areas in which there is more lithium in the water supply, but this sort of geographical epidemiology is fraught with confounding factors relating to wealth, preferences, culture, and migration. As researchers note here, lithium has both a narrow therapeutic window and only small effects on healthspan in mice.
The anti-depressant and mood stabilizing effects of lithium were discovered the mid 20th century, and administration of lithium salts is still the first-line therapy for bipolar disorders. Lithium can also ameliorate pathology in animal models of neurodegeneration, through multiple molecular mechanisms, and has been proposed as a therapy for Alzheimer's disease. Suggesting that it may have a broader therapeutic range, lithium can also extend lifespan in fission yeast, C. elegans, and Drosophila, in the last by inhibition of GSK-3 and activation of the transcription factor NRF2. Human survival across 18 Japanese municipalities correlated with increased lithium level in drinking water. These findings suggest that conserved molecular responses to lithium treatment could improve health during ageing in mammals. In this study, we therefore analysed the influence of lithium treatment on lifespan and parameters of health during ageing in mice.
To determine the concentration of lithium suitable to be administered in a longitudinal ageing study, we first tested the effects of lithium chloride (LiCl) in doses from 0.01 to 2.79 g LiCl per kg chow. C57Bl/6J mice fed with 1.05-2.79 g/kg LiCL in the diet showed lithium plasma levels between 0.4 and 0.8 mM/l. While plasma levels to 0.4 and 0.8 mM/l are well tolerated by human patients, at doses above 1.44 g LiCl/kg, we observed an obvious dose-dependent polydipsia combined with a distinct polyuria, pointing towards a significant degree of kidney toxicity.
We therefore carried out life-long lithium treatment in the range from 0.02 to 1.05 g/kg diet. Administration to both sexes at doses of 0.02 and 0.05 g/kg starting at 3 months or 18 months of age did not affect lifespan. In an additional group, treatment of females with 0.1 g/kg starting at 19 months of age also had no significant effect. Treatment of male and female mice from an age of 3 months with 0.02 and 0.05 g/kg LiCl, and then switching late in life at 22 months to 0.5 and 1.05 g/kg, respectively, had no effect on male survival and reduced maximum lifespan of females (survival of last 20% of animals to die).
We assessed the effects of lithium on other age-related phenotypes of the mice. Decreased fat mass despite unaffected food consumption indicates an effect of lithium on lipid metabolism. Mice on the low doses of 0.02 and 0.05 g/kg LiCl administered from 3 months of age showed delayed age-related loss of glucose tolerance. In addition, male mice that were switched to 0.5 and 1.05 g/kg at 22 months, after being treated with 0.02 and 0.05 g/kg from an age of 18 months, respectively, showed significantly increased tolerance to glucose at ages over 26 months. Neither treatment improved glucose tolerance in females.
There was a dose-dependent increase in motor function on the rotarod in old males under LiCl treatment, possibly related to their lower body weight. Additionally, in 24-month-old Li2CO3-treated mice, both motor function on the rotarod and endurance on the treadmill were significantly increased in males, with no effect in females. Histopathological analysis of 2-year-old, Li2CO3-treated, C3B6F1 mice showed reduced age-related pathologies in the kidneys, with significantly decreased kidney inflammation (leukocyte infiltration) in both sexes, which in males coincided strongly with a reduction of glomerulopathy.
Considering the use of a broad range of well-tolerated lithium concentrations, different lithium salts and different mouse strains, we conclude that, in contrast to the findings in yeast, worms, and flies, lithium does not seem to be a promising candidate for geroprotection in humans. Although it caused mild improvements in body weight and composition, glucose tolerance and motor performance, these were largely confined to males and were not accompanied in either sex by increased lifespan.